telephone circuit

--
If you\'re going to use the remote phone for remote control purposes
you could have it shake hands with the controller:

      LOCAL                           REMOTE

DIAL---------------------------->DETECT RING
WAIT FOR HANDSHAKE               GO OFF-HOOK
RX HANDSHAKE RX DTMF DATA

JF

OTE

OR DTMF DATA

Yes but how is this handle in circuitry. I just realize that I cannot use the same circuit with audio transformer as I did in DTMF to signal out. How should the voltage at the line while dialing? and once the party has pick up the phone ?

Ken

Yes but how is this handle in circuitry. I just realize that I cannot use the same circuit with audio transformer as I did in DTMF to signal out. How should the voltage at the line while dialing? and once the party has pick up the phone ?

Ken

Examine an old internal PC modem to see what general components are used, the telephone line is energised with about 48V, there are various examples of telephone circuitry that can be found with google, often a modem presents the AC terminals of a bridge rectifier (W04 etc) the line current is usually modulated by a high voltage transistor across the bridge rectifier DC terminals.

When the central office connects to the called phone, it sends the 20-Hz

100-volt ring signal to that phone and a tone-burst signal called the ringback to you. When someone making a call says he "hears it ringing", he's referring to the ringback. I don't remember offhand the two frequencies involved, but a little Googling will find them. Send the DTMF dial string, then wait for the ringback. Measure the time between tone bursts, because it may vary between central offices. If that much time elapses without a ringback burst, you know the other end has gone off-hook. To be safe, you also need to detect the error conditions: a busy signal or complex audio that doesn't sound like the ringback or busy tones. That will be one of the call-not-completed voice messages.

some exchanges will signal when the remote end picks up (and when they hang up) by reversing the polarity. if you don't have that you have to do what modems do and listen for a signal from the remote end. or do what fax machines do and send a tone periodically uitil you get a reply.

Bye. Jasen

That's always a problem for loop-start lines, which I presume is what you have. They're designed for calls placed by humans, to humans, rather than machine-to-machine trunks. There isn't any 100% positive electrical signal, such as a voltage level, current flow, etc. that's transmitted end-to-end & can alert a dumb machine to a distant-end answer.

What's usually used to determine answer is call progress tone detection. Call progress tones are the various tones that alert a human caller to the status of the call. Common call progress tones include dial tone, ring-back, busy, & reorder. All of these tones appear within the audio frequency band of the circuit, mostly as pairs of frequencies. If you've ever called to or from certain PBXes, or various countries, you should have noticed that these tones can vary considerably, both in terms of frequency but also cadence. In North America, there are standards for the public network, but once you've left the public network & connect to a private PBX, you may encounter something else entirely. Same for other countries, but with different standards. Depending on where you're calling FROM, as well as TO, you'll need to be able to reliably detect a variety of call progress tones.

$file/AN-CPTS-R1.pdf

JM

line/tone-search.htm

I have been reading up on them, these are call progress IC . They are now next to impossible to get. I guess now they do a DSP version so they dont need this hardware. Anybody whith an algorithm out there ?

K

--
Of course you can.

If you\'re using a proper transformer, then DTMF is entirely within the
bandwidth of the transformer.

al

n

REMOTE

T FOR DTMF DATA

well that was before I knew that Central office sends out different frequency on the status the phone of the receiver. I will look for those frequencies now.

440Hz and 480Hz means that is it ringing, if it stops, then I suppose some pickup at the other end, Then I can program the controller to send my signal. K

And whose ears are you going to fill with DTMF? If the remote end is also an automated site under your control, a DTMF handshaking routine can be set up, as has been pointed out to you. It's been done before, believe it or not. If the other end is not an automated site under your control, then you are going to very quickly become an extremely annoying person.

Well lets say I am at work, and my alarm system goes off, it will trigger the dialer. First, I will see on my phone that someone is calling me from home from the call display. Then if I hear a continuous dial tone for 20 seconds, then I will know my alarm is on.

K

On Thu, 4 Sep 2008 07:11:49 -0700 (PDT), lerameur wrote: : :Well lets say I am at work, and my alarm system goes off, it will :trigger the dialer. First, I will see on my phone that someone is :calling me from home from the call display. Then if I hear a :continuous dial tone for 20 seconds, then I will know my alarm is on. : :K

You've lost me....

Your alarm at home triggers and dials your work number. You see on your CLI display at work that the calling number is your home. Then you hear 20 seconds of dial tone so you will know your alarm is "ON"....??????? (scratches head in wonderment).

BTW, don't you mean, "you will know your home alarm has been triggered"?

Where is the 20 seconds of dial tone coming from? And why dial tone?

I suggest you start out with something like this project and refine it if necessary.

Haven't looked at your link yet, but this is an easy one. Since the human is in control of the situation on the receive end, answer the phone and press #99 or some other acknowledging code, and the auto-dialer can respond in whatever way it's been programmed. A whole series of DTMF communication may be programmed, depending on the sophistication of the auto-dialer apparatus.

Of course, the assumption remains that the receiving human knows that the alarm has triggered from the mere fact that home is calling, and presumably is not the wife asking to have a loaf of bread or jar of pickles picked up on the way home. Unless there is something tangible that gives value to the receiver communicating with the auto-dialer, there is no need for anything more than caller-ID at the receiving end.

OK, I found the other thread that you started about the PIC. It would have been very helpful in this thread if you would have stated that the auto-dial controller was very primitive and used pusle dialing. This makes DTMF out of the question, although for some reason you mentioned DTMF in your first post in this thread. What, exaclty, are the capabilities of your auto-dialer? If you just want to answer the phone and be hit with a tone, just warble 500 and 1000 Hz, you'll get the message.

:On Fri, 05 Sep 2008 02:02:12 GMT, Ross Herbert : wrote: : :>On Thu, 4 Sep 2008 07:11:49 -0700 (PDT), lerameur wrote: :>: :>:Well lets say I am at work, and my alarm system goes off, it will :>:trigger the dialer. First, I will see on my phone that someone is :>:calling me from home from the call display. Then if I hear a :>:continuous dial tone for 20 seconds, then I will know my alarm is on. :>: :>:K :>

:>

:>You've lost me.... :>

:>Your alarm at home triggers and dials your work number. You see on your CLI :>display at work that the calling number is your home. Then you hear 20 seconds :>of dial tone so you will know your alarm is "ON"....??????? (scratches head in :>wonderment). :>

:>BTW, don't you mean, "you will know your home alarm has been triggered"? :>

:>Where is the 20 seconds of dial tone coming from? And why dial tone? :>

:>I suggest you start out with something like this project and refine it if :>necessary. :>

: :Haven't looked at your link yet, but this is an easy one. Since the :human is in control of the situation on the receive end, answer the :phone and press #99 or some other acknowledging code, and the :auto-dialer can respond in whatever way it's been programmed. A whole :series of DTMF communication may be programmed, depending on the :sophistication of the auto-dialer apparatus. : :Of course, the assumption remains that the receiving human knows that :the alarm has triggered from the mere fact that home is calling, and :presumably is not the wife asking to have a loaf of bread or jar of :pickles picked up on the way home. Unless there is something :tangible that gives value to the receiver communicating with the :auto-dialer, there is no need for anything more than caller-ID at the :receiving end.

I know the human at the receive end controls the situation and that he can respond to the call using dtmf. But when he does this why would the sending end respond with dial tone? This would require the generation of 2 tones which are not part of the dtmf frequency plan. Ok, not hard to do in software, but it seems unneccesary. Why not respond with one of the standard dtmf tones instead of dial tone?

Out here dial tone is 400Hz + 425Hz afaik.

I agree that sending dial-tone is odd. I think the OP is just throwing out chum to see what bites. If you read the info on the link he posted here:

Message-ID:

it states that the auto-dialer will send a 300Hz tone for 30 seconds after placing the call. I think OP wants to expand on the idea a little, and is just guessing that sending dial tone as opposed to

300Hz is a better idea. I don't think his auto-dialer is capable of DTMF, after reading the message referenced above. Much is left unexplained about the project he is pursuing.

te:

CLI

seconds

head in

?

if

n

ng end

h are

ut it

stead

After reading on the subject I realize that the Central office sends out signals about state of the connection (ringing, busy). I am going to get a transceiver chip so I can dial out and get the signals, I think the MTC888 should be good for this. K